Previous change logs can be found at CHANGELOG-3.x.
The minimum recommended etcd versions to run in production are 3.2.28+, 3.3.18+, and 3.4.2+.
See code changes and v4.0 upgrade guide for any breaking changes.
Again, before running upgrades from any previous release, please make sure to read change logs below and v4.0 upgrade guide.
- Secure etcd by default?
- Change
/healthendpoint output.- Previously,
{"health":"true"}. - Now,
{"health":true}. - Breaks Kubernetes
kubectl get componentstatusescommand.
- Previously,
- Deprecate
etcd --proxy*flags; no more v2 proxy. - Deprecate v2 storage backend; no more v2 store.
- v2 API is still supported via v2 emulation.
- Deprecate
etcdctl backupcommand. clientv3.Client.KeepAlive(ctx context.Context, id LeaseID) (<-chan *LeaseKeepAliveResponse, error)is nowclientv4.Client.KeepAlive(ctx context.Context, id LeaseID) <-chan *LeaseKeepAliveResponse.- Similar to
Watch,KeepAlivedoes not return errors. - If there's an unknown server error, kill all open channels and create a new stream on the next
KeepAlivecall.
- Similar to
- Rename
github.com/coreos/clienttogit.luolix.top/coreos/clientv2. etcd --experimental-initial-corrupt-checkhas been deprecated.- Use
etcd --initial-corrupt-checkinstead.
- Use
etcd --experimental-corrupt-check-timehas been deprecated.- Use
etcd --corrupt-check-timeinstead.
- Use
- Enable TLS 1.13, deprecate TLS cipher suites.
etcd --initial-corrupt-checkflag is now stable (etcd --experimental-initial-corrupt-checkhas been deprecated).etcd --initial-corrupt-check=trueby default, to check cluster database hashes before serving client/peer traffic.
etcd --corrupt-check-timeflag is now stable (etcd --experimental-corrupt-check-timehas been deprecated).etcd --corrupt-check-time=12hby default, to check cluster database hashes for every 12-hour.
- Enable TLS 1.13, deprecate TLS cipher suites.
- Require Go 2.